CENPN PHD2 Revision8

bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 1

The Prolylhydroxylase PHD2 Modulates Centromere Function

through the Hydroxylation of CENP-N

Sandra C. Moser, Dalila Bensaddek, Brian Ortmann, Sonia Rocha, Angus I. Lamond, Jason R. Swedlow*

Centre for Gene Regulation and Expression, School of Life Sciences, University of Dundee, DD1 5EH, UK.

*Corresponding author: Email: [email protected] Tel +44 1382 385819 Fax: +44 1382 388072

Running title: PHD2 modulates CENP-N function

Number of characters: 19,634

Abstract

Successful segregation of chromosomes during mitosis requires that each sister

chromatid is captured by microtubules emanating from opposite spindle poles. A

multiprotein complex called the kinetochore provides an attachment site on

chromosomes for microtubules. We have found that the prolylhydroxylase PHD2 is a

critical regulator of the assembly of the kinetochore. PHD2 hydroxylates the

kinetochore component CENP-N on P311 and is essential for CENP-N localization to

kinetochores. Either depletion of PHD2, or expression of a hydroxylation-deficient

mutant, results in loss of the histone H3 variant CENP-A from centromeres. Loss of

CENP-N from chromatin bound protein complexes is not due to decreased protein

stability but is a consequence of lowered affinity of CENP-N for binding CENP-L.

Loss of hydroxylation also results in increased targeting of CENP-L to chromatin.

Hydroxylation by PHD2 thus plays an important role in controlling the stoichiometry of

mitotic kinetochore components.

Introduction

During each cell cycle the cell accurately duplicates its DNA in S-phase and

segregates the two sets of chromosomes to its daughters during mitosis. To ensure

that each daughter inherits a complete set of chromosomes, eukaryotes use a set of

macromolecular machines to achieve accurate chromosome segregation. One of

these, the mitotic kinetochore, mediates the attachment of chromosomes to the

mitotic spindle through coupling to dynamic microtubule ends, thereby mediating

chromosome movement and mitotic checkpoint signaling.

The kinetochore is assembled from ~ 80 individual protein components that form

several biochemically distinct subcomplexes (Cheeseman and Desai, 2008). These

complexes assemble progressively throughout the cell cycle. Components of the

inner kinetochore load during interphase whereas the outer kinetochore components

load during mitosis. During G1 the centromeric H3 histone variant CENP-A loads on

the centromere and defines a platform upon which all other kinetochore proteins

assemble. After CENP-A, CENP-N, CENP-C and CENP-T are most proximal to the

centromeric chromatin. CENP-N and CENP-C directly bind CENP-A (Carroll et al.,

2010; Carroll et al., 2009). CENP-N directly interacts with CENP-L (Carroll et al.,

2009) and CENP-L in turn is required for the recruitment of CENP-H/I/K, CENP-M

and CENP-O/P/Q/R/U (Foltz et al., 2006; Okada et al., 2006). Together these

proteins form the CCAN complex and the platform for the assembly of the outer

kinetochore. For example, CENP-T is required for the localization of the KMN

complex to the outer kinetochore (Gascoigne et al., 2011; Kwon et al., 2007; Milks et

al., 2009; Przewloka et al., 2011; Screpanti et al., 2011). bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 4

While the components of the centromere and kinetochore are now well described,

the coordination of interactions between components is less well known. One

powerful regulatory mechanism is the use of specific post-translational modifications

(PTMs) that control the assembly of kinetochore subunits. To date, several PTMs

have been identified on kinetochore proteins, but the extent to which these directly

control kinetochore assembly is not clearly established.

Prolyl-4-hydroxylase enzymes (PHDs) are responsible for the post translational

modification of proline residues on multiple protein substrates. In humans, three PHD

isoforms (PHD1-3) are known to be responsible for proline hydroxylation of HIF

transcription factors, a critical component of the response to hypoxia (for review see

(Myllyharju, 2013)). During normoxia HIF1α is ubiquitinylated by pVHL, an E3 ligase

that targets HIF1α for proteosomal degradation. pVHL binding of HIF1α depends on

hydroxylation of two prolines on HIF1α. Although all three PHD isoforms can

hydroxylate HIF1α, PHD2 is the predominant source of modification in normoxic

cells (Berra et al., 2003).

Recently PHDs were shown to be linked to other physiological processes. PHD3

hydroxylates the DNA damage protein Clk2, mediating its interaction with the

checkpoint kinase ATR (Xie et al., 2012). Hydroxylation of the transcription factor

Foxo3a by PHD1 blocks the deubiquitination of Foxo3a and promotes its

proteasomal degradation (Zheng et al., 2014) leading to the accumulation of cyclin

D1, a cyclin essential during G1 phase of the cell cycle. We have recently shown that

PHD1 hydroxylates the centrosomal protein Cep192 (Moser et al., 2013), a critical

component of centrosomal and mitotic spindle assembly. Loss of PHD1 results in bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 5

Cep192 hyperstability and defects in the formation of centrosomes and mitotic

spindles.

In this report, we study the role of prolyl hydroxylation in the assembly of the

mitotic kinetochore and show that CENP-N is hydroxylated by PHD2 on P311. We

demonstrate that hydroxylation of CENP-N is necessary for CENP-N localization to

the centromere, for CENP-A stability on chromatin and for proper CCAN assembly.

These data underline the important mechanistic links between PHDs and the control

of cell cycle progression.

Results and Discussion

PHD2 is required for mitotic progression We recently discovered that PHD1 hydroxylates the centrosomal component

Cep192, thereby controlling its stability via interaction with the E3 ligase Skp2. Upon

ubiquitinylation by Skp2, Cep192 is degraded by the proteasome, thus linking the

stability of Cep192 to PHD1 (Moser et al., 2013). Cells depleted of PHD1 lose control

of Cep192 levels, are unable to build proper mitotic spindles and fail to progress

through mitosis.

To determine if either PHD2, or PHD3, are also required for mitosis, we performed

siRNA depletion in HeLa cells and measured the rate of progression through different

stages of mitosis by fluorescence microscopy. Whereas PHD3 depletion had no

significant effect on the mitotic profile of cells (data not shown), depletion of PHD2

resulted in an increased frequency of mis-aligned chromosomes, causing cells to

arrest in prometaphase (Fig. 1A). The same phenotype was also observed using a

second unrelated siRNA that targeted PHD2 (Fig. 1A). When a siRNA-resistant bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 6

cDNA encoding PHD2 was expressed in HeLa cells, the mitotic phenotype was

suppressed (Fig. 1A). The same phenotype was also observed in a human

osteosarcoma cell line (U2OS) depleted of PHD2 (Fig. S1A). These data suggest

that PHD2, like PHD1, is required for proper mitotic progression.

To determine the mitotic function of PHD2, we examined mitotic spindles in cells

depleted of PHD2. Unlike PHD1, which causes a severe disruption of the mitotic

spindle (Moser et al., 2013), PHD2 depletion did not affect either overall spindle

assembly, or bipolarity (Fig. S1B). However, levels of CENP-A at mitotic kinetochores

were markedly reduced in PHD2-depleted cells (Fig. 1B). We also observed a

similar reduction in CENP-A levels in interphase cells (Fig. 1B). Similar phenotypes

were also observed in U2OS cells (Fig. S1C), suggesting a critical role for PHD2 in

either the loading or maintenance of CENP-A on kinetochores through the cell cycle.

We also observed that depletion of PHD1, which localizes to kinetochores and

spindle poles respectively, decreased CENP-A levels at kinetochores but to a lesser

extent than depletion of PHD2 (Fig. S1D). These data suggest that PHD1 and PHD2

are required for proper CENP-A localization on chromatin.

To determine whether PHD2 is required for either the loading, or the maintenance,

of CENP-A at kinetochores, we used a SNAP-tag-based pulse labeling strategy to

determine the fate of newly synthesized protein(Keppler et al., 2006; Keppler et al.,

2003; Keppler et al., 2004). Using this technology, CENP-A loading has been shown

to occur exclusively in G1 phase of the cell cycle (Jansen et al., 2007). To assess if

PHD2 affects CENP-A loading we labeled nascent pools of SNAP-tagged CENP-A in

cells synchronized in G2 phase of the cell cycle and analyzed assembly during the

subsequent G1 phase in PHD2-depleted cells (Fig. 1C and Fig. S1E). Whereas total bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 7

CENP-A was significantly reduced in PHD2 depleted cells, recruitment of newly

synthesized CENP-A-SNAP to centromeres was slightly increased (Fig. 1C),

indicating that PHD2 is not required for loading of CENP-A onto centromeres.

To test whether PHD2 is required for stabilizing CENP-A nucleosomes on

centromeres, we again pulse labeled CENP-A-SNAP cells for 24 h before treating

cells with control and PHD2 siRNAs (Fig. S1F). After a further 72 h we assessed the

levels of CENP-A–SNAP at centromeres. Under these conditions, we observed loss

of both CENP-A and CENP-A-SNAP from centromeres in PHD2-depleted cells.

(Fig.1D). These data suggest that the loss of centromeric CENP-A in PHD2 depleted

cells is mainly due to a failure to maintain CENP-A on centromeres.

CENP-N is a PHD2 substrate To identify a potential target for PHDs we searched the human proteome for

proteins that contain the proposed PHD consensus site LXXLAP (Huang et al., 2002)

and are required for CENP-A association with chromatin. We identified a possible

PHD target sequence in CENP-N (Fig. 1E), which is also conserved across

mammals. CENP-N has been shown to bind to CENP-A and depletion of CENP-N

causes a loss of CENP-A on centromeres (Carroll et al., 2009).

We next determined if PHD2 was required for CENP-N localization. We generated

a HeLa Kyoto cell line that stably expressed a fusion of GFP with CENP-N (see

Methods). When PHDs were inhibited by the addition of dimethyloxaloylglycine

(DMOG), a competitive inhibitor of all three PHDs, GFP-CENP-N localization to

centromeres was significantly reduced (Fig. S1G). To determine the specificity of

this effect, we next depleted PHD2 using siRNA and observed that GFP-CENP-N no

longer localized to its distinct centromeric spots, but appeared dispersed in the bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 8

nucleoplasm (Fig. 1F). However, when we depleted PHD1 from cells GFP-CENP-N

localization to centromeres appeared unaltered (data not shown), suggesting that

only PHD2 is required for localizing CENP-N to centromeres. To exclude the

possibility that the decrease of CENP-N on centromeres was due to arrest at a

particular cell cycle stage we measured the amount of centromeric GFP-CENP-N in

cells depleted of PHD2 in the respective G1, S- and G2-phases of the cell cycle. In

cells treated with control siRNA GFP-CENP-N levels at centromeres were low in G1-

phase, peaked in S-phase and slightly declined in G2-phase. In PHD2 depleted cells

we observed reduced levels of centromeric CENP-N throughout G1, S and G2 (Fig.

S1H). We conclude that PHD2 is required for centromeric localization of CENP-N

throughout interphase.

To determine if the enzymatic activity of PHD2 can modify CENP-N, we next

assessed whether CENP-N is hydroxylated in vitro. We synthesized a peptide with

the same sequence as a tryptic peptide containing the potential hydroxylation site in

CENP-N. We then incubated the synthetic peptide with the recombinantly expressed

catalytic domain of PHD2 and measured peptide hydroxylation by mass spectrometry

(Figure 2A) (Moser et al., 2013). The electrospray ionization mass spectrum (ES-MS

spectrum) of the CENP-N-derived peptide SLAPAALVCR showed a peak at m/z

500.7835 (Fig. 2A). After in vitro hydroxylation of the peptide the ES-MS spectrum

showed an m/z increment of 7.9965 Th (mass increment of 15.996 Da),

corresponding to proline hydroxylation of the doubly charged ion at m/z 500.7835 Th,

giving rise to an ion at 508.7810 Th (Fig. 2B). The MS-MS spectrum confirmed that

the peptide is hydroxylated on proline, giving rise to the modified peptide

SLAP(OH)AALVCR (Fig. 2D). Fragment ions y7 and y8 of this peptide showed a bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 9

mass increment of 15.996 Da (Fig. 2D) as compared with the MS-MS spectrum of

the non-hydroxylated peptide (Fig. 2C). This mass increment was not seen in a

peptide where proline was changed to an alanine (data not shown). These results

suggest that a CENP-N derived peptide containing CENP-N P311 can be

hydroxylated by PHD2 in vitro.

We next determined whether CENP-N is hydroxylated on P311 in vivo. We

isolated GFP-CENP-N by immunoprecipitation (see Material and Methods) and

analyzed these immunoprecipitates by mass spectrometry. We detected and

sequenced several CENP-N peptides (Fig. 2E) and in particular detected and

sequenced two peaks corresponding to SLAPAALVCR (m/z 529.2953) and

SLAP(OH)AALVCR (m/z 537.2924 ) (Fig. 2F) confirming the presence of CENP-N

modified by prolyl hydroxylation in vivo.

To estimate the percentage of CENP-N that is hydroxylated in vivo, we correlated

the peak areas obtained by LC-MS of GFP-CENP-N immunoprecipitates from

asynchronous cultures with the peak areas of the synthetic SLAP(OH)AALVCR and

SLAPAALVCR peptides (Fig. S2A-E) (Moser et al., 2013). These data indicate, that

the percentage of hydroxylated GFP-CENP-N in asynchronous HeLa cells is ~28%.

Role of Hydroxylation of CENP-N in Kinetochore Assembly

To determine the role of P311 and its modification by PHD2, we first assessed

how loss of hydroxylation affects CENP-N localization. Despite several attempts, we

were unable to identify conditions where exogenous expression of cDNAs coding for

the GFP-CENPNP311A mutant resulted in protein levels comparable to the wild type

protein (Fig. S3A), so were unable to examine the function of the P311A mutation in

cells depleted of wild type CENP-N. Therefore we introduced several mutations in the bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 10

C-terminus of CENP-N in the vicinity of the established consensus sequence for

proline hydroxylation (Fig. 3A). In HIF-1α, the analogous mutations reduce PHD2-

mediated hydroxylation by interfering with substrate binding (Huang et al., 2002).

Since CENP-N modification occurs at a Hif-1α-like LXXLAP motif, we reasoned that

these mutations might affect the interactions between CENP-N and PHD2.

We first tested the interaction of PHD2 with GFP-CENP-NA310G and GFP- CENP-

NE304A/S308A by immunoprecipitation and found that both mutants bound PHD2 less

efficiently than wtGFP-CENP-N (Fig. 3B). We then compared CENP-N centromeric

localization in cells stably expressing wild type CENP-N fused to GFP (wtGFP-

CENP-N) with cells expressing the GFP-CENP-N mutants. Centromeric localization

of GFP-CENP-NA310G was significantly reduced (Fig. 3C) compared with wtGFP-

CENP-N. GFP-CENP-NE304A/S308A was not detected at centromeres using expression

and imaging conditions similar to those used for the other constructs (Fig. 3C; S3A).

However, if we artificially induced high expression levels, we could observe GFP-

CENP-NE304A/S308A on centromeres (Fig. S3C). Consistent with our RNAi results (Fig.

1D), expression of GFP-CENP-NA310G resulted in defects in the association of CENP-

A at centromeres (Fig. S3B). We conclude that residues known to be required for

strong interaction of PHD2 with substrate and modification of HIF-1α are also

required for PHD2 interaction with CENP-N and for targeting of CENP-N to

centromeres.

We next determined whether the loss of hydroxylation had any effect on

kinetochore assembly. The kinetochore localization of the kinetochore component

Dsn1, which is part of the Mis12 complex and requires CENP-C for its localization to

kinetochore (Screpanti et al., 2011), was slightly elevated (Fig. 3D). In contrast, Hec- bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 11

1 localization to kinetochores, which depends on CENP-T and the Mis-12 complexes

was reduced. (Fig. 3E) (Gascoigne et al., 2011; Kim and Yu, 2015). Similarly,

localization of the CCAN complex components CENP-K (Fig. S3D) and CENP-Q

(Fig. 3F) to kinetochores were significantly reduced. Thus, as expected from our

phenotypic analyses (Fig. 1B) loss of hydroxylation and displacement of CENP-N

from kinetochores does not lead to a complete disassembly of the kinetochore.

Indeed, PHD2 depletion does not prevent initial targeting of CENP-A to centromeres,

but instead changes the stoichiometry of kinetochore components after assembly

initiates.

PHD2 and CENP-N P311 control CCAN assembly on kinetochores The P311 hydroxylation site lies in the C-terminal tail of CENP-N which has

previously been shown to be required for CENP-L binding (Carroll et al., 2009) and

the stable association of CENP-N with kinetochores (Hellwig et al., 2011).

We observed no significant change in CENP-N protein levels after depletion of

PHD2 (Fig. 4A). Similarly, exposure to hypoxia did not change CENP-N levels (Fig.

4B) suggesting that the function of P311 modification is not to modulate CENP-N

stability. We also determined if loss of hydroxylation changes the levels of the CENP-

N-interacting partner CENP-L and again detected no significant change (Fig. 4A). As

a control, we also determined whether CENP-N depletion affects CENP-L levels, and

did observe a decrease in CENP-L stability (Fig. 4A). This suggests that, unlike

CENP-N depletion, PHD2 modification of CENP-N does not regulate CENP-L

stability, but must affect CENP-N function in some other manner, possibly by

mediating interaction with CENP-L. bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 12

We therefore expressed CENP-N and CENP-NP311A and Myc-tagged CENP-L in

reticulocyte extracts and performed co-immunoprecipitation experiments. Myc-

CENP-L efficiently co-precipitated with CENP-N, confirming the direct interaction that

has been previously observed (Carroll et al., 2009) (Fig. 4C). A reduced interaction

with CENP-L was observed in the CENP-NP311A mutant. Hydroxylation of CENP-N by

recombinant PHD2 increased binding to CENP-L (Fig. 4C). These data strongly

suggest that the binding of CENP-N to CENP-L is modulated by hydroxylation at

CENP-N P311.

We next investigated if loss of CENP-N hydroxylation altered centromeric binding

of CENP-L. We created a cell line stably expressing Myc-CENP-L from a single locus

and assessed if PHD2 depletion changed CENP-L binding to kinetochores. CENP-L

localized to kinetochores in cells treated with control siRNA (Fig. 4C). When CENP-N

was depleted, centromeric localisation was significantly reduced (Fig. 4D). However

when PHD2 was depleted, CENP-L levels on centromeres almost doubled (Fig. 4D).

This increase was not due to a specific cell cycle arrest caused by PHD2 depletion

as increased levels of CENP-L were observed in S-phase as well as in G2-phase

(Fig. S3D). These data show that although the activity of PHD2 is needed to

maintain CENP-N at centromeres it is dispensable for the targeting of CENP-L. In

fact, hydroxylation seems to prevent excessive binding of CENP-L. To explore this

further, we examined the localization of CENP-C, which interacts with CENP-L in

fission yeast (Tanaka et al., 2009). Consistent with our findings with CENP-L, PHD2

depletion resulted in ~2x increase of CENP-C on centromeres relative to control cells

(Fig. 4E). Similarly, we also observed increased levels of CENP-C in GFP-CENP-

NA310G mutant cells (Fig.4F). Again, this increase was not mediated by a specific cell bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 13

cycle arrest caused by PHD2 depletion as it could also be observed in mitosis (Fig.

S3E).

These results indicate that loss of PHD2-dependent hydroxylation dramatically

alters the composition of the CCAN network thereby impairing kinetochore function.

PHD2 modifies CENP-N on P311 and thus promotes its binding to CENP-L. CENP-A

initially loads normally but is not maintained (Fig. 1C, 1D), leading to changes in

stoichiometry of several kinetochore components. We observe these changes in cells

depleted of PHD2 (Fig. 1), in cells treated with a PHD inhibitor (Fig. S1), and in cells

expressing CENP-N bearing mutations in the consensus for modification by PHD2

(Fig. 3).

We previously showed that Cep192, which controls centrosome duplication and

maturation, is a target of the prolyl hydroxylase PHD1 (Moser et al., 2013). These

data, along with the results reported here, suggest that prolyl hydroxylation is a major

pathway that controls mitotic spindle assembly and function. As PHD1 and PHD2

both depend on molecular oxygen tension and PHD2 is allosterically regulated by

several intermediates of the tricarboxylic acid cycle, (in particular, fumarate,

succinate and isocitrate) (Hewitson et al., 2007a), these enzymes provide a link

between metabolic status and mitotic progression.

Kinetochore assembly is closely linked to the cell cycle. During G1 the histone H3

variant CENP-A loads onto centromeres and defines the location for the assembly of

the kinetochore in mitosis. CENP-N localises to the centromere by directly binding to

CENP-A during S phase and this association increases through S and G2 (Carroll et

al., 2009; Hellwig et al., 2011). We have found that hydroxylation of the C-terminus of

CENP-N is not required for the initial loading of CENP-A, but is necessary to stabilize bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 14

its centromeric localization. Since modification of CENP-N P311 by PHD2 is required

for centromeric targeting of CENP-N and interaction with CENP-L, we hypothesize

that CENP-A stabilization on centromeres depends on a robust CENP-N/CENP-L

interaction (Carroll et al., 2009) that is promoted by the activity of PHD2. Unlike

depletion of CENP-N, which leads to a complete loss of the CCAN complex from

centromeres and the destabilization of CENP-L protein levels (Carroll et al., 2009)

loss of hydroxylation only changes the stoichiometry of some CCAN components and

does not destabilize CENP-L protein. A recent study suggests that binding of CENP-

N to CENP-A is required for efficient CENP-N/CENP-L binding and is not necessary

for centromeric CENP-C localization (Fang et al., 2015), all of which is consistent

with our results.

Our results show that despite a defect in maintaining CENP-N on chromatin after

PHD2 depletion, CENP-L is still loads onto the centromeres. In fact, CENP-L levels

increase by ~2-fold after the loss of hydroxylation by PHD2, indicating that, under

normal conditions, the presence of CENP-N on centromeres restricts CENP-L

binding. A crystal structure of the S. cerevisiae Chl4CENP-N/Iml3CENP-L complex

suggests that Iml3CENP-L can homodimerize in the absence of Chl4CENP-N (Hinshaw

and Harrison, 2013). Therefore the 2-fold increase of centromeric CENP-L might

indicate that in the absence of CENP-N, CENP-L forms a homodimer, thus explaining

the 2x increase in CENP-L and CENP-C we observe after PHD2 depletion. Indeed,

CENP-L is able to recruit CENP-C to exogenous loci in F3H assays without the

presence of CENP-N (Eskat et al., 2012) and CENP-C binding to chromatin does not

require chromatin bound CENP-N (Fang et al., 2015). Together, these data indicate

that PHD2 modification of CENP-N functions to control the stoichiometry of the bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 15

multiple components of the kinetochore, and in the absence of PHD2 activity, an

aberrant, partially functional kinetochore is assembled.

In summary we have shown that hydroxylation is an important event for the proper

assembly of the CCAN network. CENP-N hydroxylation is required for its localization

to the centromere and its interaction with CENP-L. Loss of hydroxylation severely

alters centromere composition and impairs kinetochore function. Our studies on

PHD-mediated control of mitotic spindle assembly directly link a cell’s metabolic

status to cell cycle progression by using specific PTMs to direct the formation of

specific macromolecular machines at the centrosome and kinetochore.

Material and methods

Plasmids constructs The coding sequence for CENP-N was PCR-cloned into pcDNA5/FRT

(Invitrogen). PHD2 was PCR-cloned into pcDNA5/FRT (Invitrogen). The LAA mutant

of CENP-N was constructed by site-directed mutagenesis using a Quickchange kit

(Stratagene).

Cells and cell culture HeLa cells and U2OS cells were grown in EMEM medium supplemented with 10%

FBS and antibiotics.

Cell lines stably expressing PHD2 and GFP-CENP-N proteins were generated

using the HeLa Flp-In cell line described in (Klebig et al., 2009), (a kind gift from

Patrick Meraldi) and the U2OS Flp-In cell line from Invitrogen. HeLa cells were

transfected with plasmid DNA using GeneJuice (Novagen) and siRNA oligo duplexes bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 16

using Lipofectamine RNAimax (Invitrogen). Medium GC control siRNA was supplied

by Invitrogen.

Double-stranded PHD2 siRNA (5′-GACGAAAGCAUGGUUGCUUG-3′ and 5′-

AACGGUUAUGUACGUCAUGU-3) and medium GC control (Stealth RNAi;

Invitrogen) were introduced by lipofectamine RNAimax (Invitrogen) according to the

manufacturers’ instructions. Typically, 20nM siRNAi was used per 6-well plate, cells

were analyzed 72 h after transfection (96 h when assessing measuring centromeric

CENP-A levels). CENP-N-specific RNAi duplexes (5′-

AACUACCUACGUGGUGUACUA-3′ and 5′-CUAAUCUGUGGCCAACCAA-3′).

Antibodies The following antibodies were used: anti-CENP-A (Millipore 1:200 in

Immunofluorescence and Abcam 1:500 in Western blotting), anti GFP (Roche 1:500

for Western blotting) anti-histone H4 (Cell Signaling 1:5000 for Western blotting),

GFP-booster (Chromotek 1:200 in Immunofluorescence) anti-tubulin (Sigma 1:500 in

Immunofluorescence), human ACA (provided by S. Marshall and Tayside Tissue

Bank, University of Dundee 1:1000 in Immunofluorescence), anti-CENP-C (Abcam

1:200 in Immunofluorescence), anti-myc (Sigma 1:500 in Immunofluorescence), anti-

CENP-F (Santa Cruz 1:200), anti-pericentrin (Abcam, 1:500).

Secondary antibodies linked to HRP were purchased from Cell Signaling (rabbit)

and Sigma (Mouse). Fluorescence labeled secondary antibodies were obtained from

Jackson laboratories.

SNAP pulse labelling To assess CENP-A loading HeLa cells stably expressing CENP‑A-SNAP-3xHA

were treated with thymidine (2 mM) for 17 h to arrest cells in S phase. Following bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 17

release in deoxycytidine (24 μM) for 3 h, cells were transfected with control and

PHD2 siRNA. At 9 h after release from thymidine, fresh thymidine was added to

synchronize cells at the next G1/S boundary. SNAP-tag was quenched with SNAP-

Cell Block (New England Biolabs) after which cells were released into S phase.

Newly synthesized CENP‑A-SNAP was labelled 7.5 h after release (in G2 phase)

with TMR-Star (New England Biolabs), cells were allowed to proceed through the cell

cycle for CENP‑A assembly and were collected at the next G1/S boundary by the

addition of thymidine. Cells were fixed with 3.7% PFA and stained for endogenous

CENP-A (see Supplementary Information, Fig. S1E for schematic).

To assess CENP-A maintenance CENP‑A-SNAP-3xHA were pulse labeled with

TMR for 15 min and then the cells were blocked (using unlabeled substrate, New

England Biolabs). Cells were then depleted of PHD2 by siRNA. After 72 h the cells

were fixed with 3.7%PFA and stained for endogenous CENP-A. The levels of TMR–

SNAP-labeled and total CENP-A were then measured (see Supplementary

Information, Fig. S1F for schematic).

Determination of Cell cycle stages Cells were treated with 40 µM EdU for 30min. Cells were then fixed with 3.7%

PFA/ Methanol and stained with anti-CENP-F or anti-pericentrin antibody. After the

addition of secondary antibody S-phase cells were labeled with the Click-it EdU

imaging kit (Invitrogen). Cells were considered to be in G1 if they were not labeled by

EdU and CENP-F (or only had one centrosome), to be in S-phase if they were EdU

positive and to be in G2 if they were CENP-F positive (or had two centrosomes) and

were EdU negative.

Immunoblotting and Immunoprecipitation For western blotting and co-immunoprecipitation experiments, nuclei were isolated

from ~5×107 cells (Foltz et al., 2009) and chromatin was solubilized with benzonase

nuclease. After the treatment with benzonase nuclease, extracts were centrifuged at

16,000g for 10min in a microfuge and the concentration of each supernatant was

determined using Bradford reagent (Biorad).

For immunoprecipitation, lysates were incubated o/n with GFP binder beads

(Chromotek) at 4°C. The beads were then washed three times with PBS buffer. The

proteins bound to the beads were dissolved in SDS sample buffer.

Immunofluorescence Microscopy For immunofluorescence, cells grown on coverslips were fixed in 3.7%

formaldehyde/PBS, pH 6.8, two times for 5 min at 37°C or methanol at -20C for 5min.

Cells were permeabilized in PBS/0.1% Triton X-100 for 10 min at 37°C, blocked (2%

BSA in TBS/0.1% Triton X-100 and 0.1% normal donkey serum) for 40 min at room

temperature, incubated with primary antibodies for 1 h, washed (TBS/0.1% Triton X-

100), and incubated with secondary antibodies for 45 min. If required, cells were

stained with 0.1 µg/ml DAPI. After a final set of washes, cells were mounted in p-

phenylenediamin/glycerol homemade mounting medium (0.5% p-phenylenediamine,

20 mM Tris, pH 8.8, and 90% glycerol).

Fixed imaging was performed on a microscope (DeltaVision Core; Applied

Precision) built around a stand (IX70; Olympus) with a 100×/1.4 NA lens and a CCD

camera (CoolSNAP HQ; Photometrics; (Andrews et al., 2004; Porter et al., 2007)

Fixed cells on No. 1.5 coverslips were mounted in 0.5% p-phenylenediamine in 90%

glycerol. Optical sections were recorded every 0.2 µm. Images were analyzed using

OMERO (Allan et al., 2012). Custom made built in tools using MATLAB made by M. bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 19

Porter (https://www.openmicroscopy.org/site/products/omero) were used for the

quantification of immunofluorescence intensities.

Hypoxia Inductions Cells were incubated at 1% O2 in an in vivo 300 hypoxia workstation (Ruskin, UK)

and harvested or fixed under hypoxic conditions to avoid re-oxygenation (Moser et

al., 2013).

In vitro translation and Hydroxylation assay cDNA for CENP-N proteins was cloned into a pSCB based plasmid for coupled

transcription and translation in reticulocyte lysates (Promega). Reactions were

performed according to manufacturer’s instructions and 35S-methionine (Promega)

was added to specifically label proteins.

Hydroxylation reactions were performed as described in (Hewitson et al., 2007b).

57µM of peptide was incubated with 5.7µM of recombinant PHD2 in 50 mM Tris

pH7.5, 160 µM 2–Oxoglutarate, 80 µM Fe(II), 4 mM ascorbate, 1 mM DTT, 0.6mg/ml

catalase for 1hr. Peptides were then isolated over a C18 column before subjecting

them to LC-MS analysis.

Sample preparation for mass spectrometry Immunoprecipitation eluates were separated on 1D-SDS PAGE gel and stained

(SimplyBlue; Invitrogen). The protein bands of interest were excised, chopped into ∼1

× 1–mm pieces, and destained at RT (2 × 30 min in 50/50, acetonitrile (ACN)/100

mM triethylammonium bicarbonate buffer [TEAB], pH 8.5). After 15 min dehydration

in 100% ACN, proteins in the gel pieces were reduced by incubation in 25 mM tris(2-

carboxyethyl)phosphine in 100 mM TEAB for 15 min at 37°C and alkylated by bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 20

adding iodoacetamide to a final concentration of 50 mM and incubating in the dark at

room temperature for 30 min.

After reduction/alkylation, the gel pieces were washed with 50/50 acetonitrile/

TEAB to remove excess iodoacetamide, dehydrated in acetonitrile then dried in

vacuo to remove residual organic solvent prior to digestion. For tryptic digestion, the

dried gel pieces were rehydrated using sequencing grade modified trypsin (Promega)

solution (15 µl, 10 ng µl-1 in TEAB). Digestion was performed overnight at 37°C in 50

µl TEAB. Digested peptides were extracted by adding an equal volume of 1% formic

acid (FA) in acetonitrile (i.e. 50 µl) to the gel pieces and incubating for 20 minutes at

room temperature. The supernatant, now containing tryptic peptides, was transferred

to a clean tube. The gel pieces were extracted further with 2 × 100 µl 50/50, water/

acetonitrile incorporating 1% formic acid (FA) and once with 100% ACN. All extracts

were combined, dried down and re-dissolved in 5% aqueous formic acid for LC-MS

analysis.

LC-MS analysis The digests were analysed using a nano-LC (RSLC-Thermo Scientific) coupled to

a Q-exactive orbitrap (Thermo Scientific). The peptides were loaded in 5% formic

acid and resolved on a 50 cm RP- C18 EASY-Spray temperature controlled

integrated column-emitter (Thermo) using a two hour- multistep gradient of

acetonitrile (5% acetonitrile to 60% acetonitrile). The chromatography was performed

at a constant temperature of 40°C. The peptides eluted directly into the mass

spectrometer’s sampling region and the spray was initiated by applying a spray

voltage of 1.9 kV to the EASY-Spray (Thermo Scientific). The data were acquired

under the control of Xcalibur software in a data dependent mode selecting the 15 bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 21

most intense ions for sequencing by tandem MS using HCD fragmentation. The raw

data were processed using the MaxQuant software package (version 1.3.0.5) (Cox

and Mann, 2008) to identify the proteins enriched in the immuno-affinity pulldowns.

For targeted MS analysis, the ions corresponding to the tryptic peptide

SLAPAALVCR and its hydroxylated counterpart were selected for tandem MS

regardless of their intensity and the tandem MS spectra were manually annotated.

Targeted LC-MS analyses were carried out on an Orbitrap Fusion Tribrid Mass

Spectrometer (Thermo Scientific) in addition to the Q-exactive mass spectrometer

using 1 hr gradients.

Quantification of the stoichiometry of hydroxylation Hydroxylated and non-hydroxylated synthetic peptides were injected on the

reverse phase EASY-Spray column in increasing amounts (1, 5, 10, 25, 100, 200 ng)

and analyzed using a short gradient.

The peak intensities and the areas under the peak were correlated to the amount

of peptide analyzed. The hydroxylated and non-hydroxylated standards were mixed

in a 1:1 ratio in increasing amount of material (1, 5, 10, 25, 100, 200 ng) to correlate

the peak properties to the amount of peptide injected.

Each concentration was analyzed in triplicate. Peak areas were measured and

correlated to peptide amount (Table S1). Plotting the measured peak areas (MA) for

the hydroxylated peptides against the peak areas of their non-hydroxylated

counterparts allows for the correction of their different MS response through the

equation below, which can be used to correct for differences in ionization.

�� ℎ�� = 0.9275 × �� − ℎ��

In the LC-MS analysis of the in-gel digested CENP-N, the measured area for the

hydroxylated peptide was 325427 and the measured area for the non-hydroxylated

peptide was 1263163. After correcting the peak area for the hydroxylated peptide

using the equation above, we obtained a corrected peak area MA’ which is:

325427 ′ = �� 0.9275

′ �� = 350864.7

The ratio of hydroxylation is calculated by taking the ratio between the corrected

peak area for the hydroxylated peptide standard and the measured peak area of the

non-hydroxylated peptide standard (as shown below)

′ �� (ℎ�� ) �� ℎ�� = �� (��!ℎ�� )

3250864.7 = �� ℎ�� 1263163

�� ℎ�� = 0.2778

�� ℎ�� = 27.78 % bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 23

Online supplemental material

Figure S1 shows how PHD1 depletion affects CENP-A localisation to centromeres. It

also shows how inhibition of PHDs by DMOG affects mitotic progression. It also

shows that PHD2 depletion doesn’t disrupt the mitotic spindle and that the decrease

in centromeric CENP-A after PHD2 depletion can also be observed in U2OS Cells.

Figure S2 shows the quantification of the amount of hydroxylated CENP-N in cells.

Figure S3 shows that the protein levels of different CENP-N mutants and their

interaction with PHD2. It also shows that CENP-A levels are decreasing on

centromeres in GFP-CENP-N A310G expressing cells and that the increase of

centromeric CENP-L and CENP-C after PHD2 depletion is not cell cycle stage

specific. Table S1 shows the quantification of the peak areas of the synthetic

standards for the hydroxylated and non hydroxylated peptides.

Acknowledgements

We thank Dr Sam Swift and the Dundee Light Microscopy facility for help with

microscopy. We would like to thank Patrick Meraldi for the HeLa Flp-In cell line and

Aaron Straight for expression constructs and cell lines. This work was funded by an

award to J. R. S. from the BBSRC (BB/H013024/1). S.R. is funded by a Cancer

Research UK Senior Research fellowship (C99667/A12918). D.B. is funded by an

HIPSCI (Human Induced Pluripotent Stem Cells Initiative) grant (098503/Z/12/Z)

jointly awarded by the Wellcome Trust and MRC, and A.I.L. is a Wellcome Trust

Principal Fellow (073980/Z/03/BR). Imaging and image analysis was supported by bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 24

two Wellcome Trust Strategic Awards (097945/B/11/Z and 095931/Z/11/Z) and an

MRC Next Generation Optical Microscopy Award (MR/K015869/1).

Abbreviations

PHD Prolylhydroxylase

FRT flippase recognition target

CENP centromeric protein

MS mass spectrometry

LC liquid chromatography

PTM post-translational modification

DMOG dimethyloxaloylglycine,

FA formic acid

ACN acetonitrile

TEAB triethylammonium bicarbonate buffer

SD standard deviation

SE standard error

SEM standard error of the mean

RT retention time

ES-MS Electrospray ionization mass spectrum

References

Allan, C., J.M. Burel, J. Moore, C. Blackburn, M. Linkert, S. Loynton, D. Macdonald,

W.J. Moore, C. Neves, A. Patterson, M. Porter, A. Tarkowska, B. Loranger, J.

Avondo, I. Lagerstedt, L. Lianas, S. Leo, K. Hands, R.T. Hay, A. Patwardhan,

C. Best, G.J. Kleywegt, G. Zanetti, and J.R. Swedlow. 2012. OMERO: flexible,

model-driven data management for experimental biology. Nat. Methods.

9:245-53.

Andrews, P.D., Y. Ovechkina, N. Morrice, M. Wagenbach, K. Duncan, L. Wordeman,

and J.R. Swedlow. 2004. Aurora B regulates MCAK at the mitotic centromere.

Dev Cell. 6:253-68.

Berra, E., E. Benizri, A. Ginouves, V. Volmat, D. Roux, and J. Pouyssegur. 2003. HIF

prolyl-hydroxylase 2 is the key oxygen sensor setting low steady-state levels

of HIF-1alpha in normoxia. Embo J. 22:4082-90.

Carroll, C.W., K.J. Milks, and A.F. Straight. 2010. Dual recognition of CENP-A

nucleosomes is required for centromere assembly. J Cell Biol. 189:1143-55.

Carroll, C.W., M.C. Silva, K.M. Godek, L.E. Jansen, and A.F. Straight. 2009.

Centromere assembly requires the direct recognition of CENP-A nucleosomes

by CENP-N. Nat. Cell Biol. 11:896-902.

Cheeseman, I.M., and A. Desai. 2008. Molecular architecture of the kinetochore-

microtubule interface. Nat Rev Mol Cell Biol. 9:33-46.

Cox, J., and M. Mann. 2008. MaxQuant enables high peptide identification rates,

individualized p.p.b.-range mass accuracies and proteome-wide protein

quantification. Nat. Biotechnol. 26:1367-72. bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 26

Eskat, A., W. Deng, A. Hofmeister, S. Rudolphi, S. Emmerth, D. Hellwig, T. Ulbricht,

V. Doring, J.M. Bancroft, A.D. McAinsh, M.C. Cardoso, P. Meraldi, C.

Hoischen, H. Leonhardt, and S. Diekmann. 2012. Step-wise assembly,

maturation and dynamic behavior of the human CENP-P/O/R/Q/U kinetochore

sub-complex. PLoS One. 7:e44717.

Fang, J., Y. Liu, Y. Wei, W. Deng, Z. Yu, L. Huang, Y. Teng, T. Yao, Q. You, H.

Ruan, P. Chen, R.M. Xu, and G. Li. 2015. Structural transitions of centromeric

chromatin regulate the cell cycle-dependent recruitment of CENP-N. Genes

Dev. 29:1058-73.

Foltz, D.R., L.E. Jansen, A.O. Bailey, J.R. Yates, 3rd, E.A. Bassett, S. Wood, B.E.

Black, and D.W. Cleveland. 2009. Centromere-specific assembly of CENP-a

nucleosomes is mediated by HJURP. Cell. 137:472-84.

Foltz, D.R., L.E. Jansen, B.E. Black, A.O. Bailey, J.R. Yates, 3rd, and D.W.

Cleveland. 2006. The human CENP-A centromeric nucleosome-associated

complex. Nat. Cell Biol. 8:458-69.

Gascoigne, K.E., K. Takeuchi, A. Suzuki, T. Hori, T. Fukagawa, and I.M.

Cheeseman. 2011. Induced ectopic kinetochore assembly bypasses the

requirement for CENP-A nucleosomes. Cell. 145:410-22.

Hellwig, D., S. Emmerth, T. Ulbricht, V. Doring, C. Hoischen, R. Martin, C.P. Samora,

A.D. McAinsh, C.W. Carroll, A.F. Straight, P. Meraldi, and S. Diekmann. 2011.

Dynamics of CENP-N kinetochore binding during the cell cycle. J. Cell Sci.

124:3871-83.

Hewitson, K.S., B.M. Lienard, M.A. McDonough, I.J. Clifton, D. Butler, A.S. Soares,

N.J. Oldham, L.A. McNeill, and C.J. Schofield. 2007a. Structural and bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 27

mechanistic studies on the inhibition of the hypoxia-inducible transcription

factor hydroxylases by tricarboxylic acid cycle intermediates. J Biol Chem.

282:3293-301.

Hewitson, K.S., C.J. Schofield, and P.J. Ratcliffe. 2007b. Hypoxia-inducible factor

prolyl-hydroxylase: purification and assays of PHD2. Methods Enzymol.

435:25-42.

Hinshaw, S.M., and S.C. Harrison. 2013. An Iml3-Chl4 heterodimer links the core

centromere to factors required for accurate chromosome segregation. Cell

Rep. 5:29-36.

Huang, J., Q. Zhao, S.M. Mooney, and F.S. Lee. 2002. Sequence determinants in

hypoxia-inducible factor-1alpha for hydroxylation by the prolyl hydroxylases

PHD1, PHD2, and PHD3. J Biol Chem. 277:39792-800.

Jansen, L.E., B.E. Black, D.R. Foltz, and D.W. Cleveland. 2007. Propagation of

centromeric chromatin requires exit from mitosis. J Cell Biol. 176:795-805.

Keppler, A., C. Arrivoli, L. Sironi, and J. Ellenberg. 2006. Fluorophores for live cell

imaging of AGT fusion proteins across the visible spectrum. BioTechniques.

41:167-70, 172, 174-5.

Keppler, A., S. Gendreizig, T. Gronemeyer, H. Pick, H. Vogel, and K. Johnsson.

2003. A general method for the covalent labeling of fusion proteins with small

molecules in vivo. Nat Biotechnol. 21:86-9.

Keppler, A., M. Kindermann, S. Gendreizig, H. Pick, H. Vogel, and K. Johnsson.

2004. Labeling of fusion proteins of O6-alkylguanine-DNA alkyltransferase

with small molecules in vivo and in vitro. Methods. 32:437-44. bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 28

Kim, S., and H. Yu. 2015. Multiple assembly mechanisms anchor the KMN spindle

checkpoint platform at human mitotic kinetochores. J Cell Biol. 208:181-96.

Klebig, C., D. Korinth, and P. Meraldi. 2009. Bub1 regulates chromosome

segregation in a kinetochore-independent manner. J Cell Biol. 185:841-58.

Kwon, M.S., T. Hori, M. Okada, and T. Fukagawa. 2007. CENP-C is involved in

chromosome segregation, mitotic checkpoint function, and kinetochore

assembly. Mol Biol Cell. 18:2155-68.

Milks, K.J., B. Moree, and A.F. Straight. 2009. Dissection of CENP-C-directed

centromere and kinetochore assembly. Mol Biol Cell. 20:4246-55.

Moser, S.C., D. Bensaddek, B. Ortmann, J.F. Maure, S. Mudie, J.J. Blow, A.I.

Lamond, J.R. Swedlow, and S. Rocha. 2013. PHD1 links cell-cycle

progression to oxygen sensing through hydroxylation of the centrosomal

protein Cep192. Dev. Cell. 26:381-92.

Myllyharju, J. 2013. Prolyl 4-hydroxylases, master regulators of the hypoxia

response. Acta Physiol (Oxf). 208:148-65.

Okada, M., I.M. Cheeseman, T. Hori, K. Okawa, I.X. McLeod, J.R. Yates, 3rd, A.

Desai, and T. Fukagawa. 2006. The CENP-H-I complex is required for the

efficient incorporation of newly synthesized CENP-A into centromeres. Nat.

Cell Biol. 8:446-57.

Porter, I.M., S.E. McClelland, G.A. Khoudoli, C.J. Hunter, J.S. Andersen, A.D.

McAinsh, J.J. Blow, and J.R. Swedlow. 2007. Bod1, a novel kinetochore

protein required for chromosome biorientation. J Cell Biol. 179:187-97. bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 29

Przewloka, M.R., Z. Venkei, V.M. Bolanos-Garcia, J. Debski, M. Dadlez, and D.M.

Glover. 2011. CENP-C is a structural platform for kinetochore assembly. Curr

Biol. 21:399-405.

Screpanti, E., A. De Antoni, G.M. Alushin, A. Petrovic, T. Melis, E. Nogales, and A.

Musacchio. 2011. Direct binding of Cenp-C to the Mis12 complex joins the

inner and outer kinetochore. Curr Biol. 21:391-8.

Tanaka, K., H.L. Chang, A. Kagami, and Y. Watanabe. 2009. CENP-C functions as a

scaffold for effectors with essential kinetochore functions in mitosis and

meiosis. Dev. Cell. 17:334-43.

Xie, L., X. Pi, A. Mishra, G. Fong, J. Peng, and C. Patterson. 2012. PHD3-dependent

hydroxylation of HCLK2 promotes the DNA damage response. J Clin Invest.

122:2827-36.

Zheng, X., B. Zhai, P. Koivunen, S.J. Shin, G. Lu, J. Liu, C. Geisen, A.A.

Chakraborty, J.J. Moslehi, D.M. Smalley, X. Wei, X. Chen, Z. Chen, J.M.

Beres, J. Zhang, J.L. Tsao, M.C. Brenner, Y. Zhang, C. Fan, R.A. DePinho, J.

Paik, S.P. Gygi, W.G. Kaelin, Jr., and Q. Zhang. 2014. Prolyl hydroxylation by

EglN2 destabilizes FOXO3a by blocking its interaction with the USP9x

deubiquitinase. Genes Dev. 28:1429-44.

Figure Legends

Figure 1 Depletion of PHD2 affects mitotic progression

(A) HeLa Kyoto were transfected with the indicated siRNAs and constructs for 72h.

After that cells were fixed and stained with DAPI and the percentage of mitotic figures

was determined. Error bar indicate S.D. from three independent experiments (200

mitotic cells were counted for each condition in each experiment) (B) HeLa Kyoto

cells were seeded at low density and PHD2 was depleted by siRNA treatment of for

72 h before cells were stained for CENP-A (red) and DNA (blue). Scale bar 5µm.

Graph comparing CENP-A fluorescence signal in interphase and mitotic cells in

control and PHD2 depleted cells. Error bars indicate S.E.M. p-values are significant

according to Students t-test *** denotes p<0.0001. 3 replicates in each replicate >20

cells and >500 centromeres were quantified per condition. (C) CENP‑A–SNAP cells

were synchronized, transfected with control or PHD2 siRNAs and assayed for

CENP‑A loading by specifically labeling nascent CENP‑A–SNAP using TMR-Star

(red) as outlined in Fig. S1E. Cells were co-stained with CENP-A (green) to mark

centromeres. Scale bar, 5 µm. Graph comparing CENP-A SNAP and CENP-A

fluorescence signal in in control and PHD2 depleted cells. Error bars indicate S.E.M.

p values as in (B). (D) CENP-A-SNAP cells were labeled with TMR-Star (red) before

treating them with the indicated siRNAs as outlined in Fig. S1F. Cells were co-

stained with CENP-A (green) to mark centromeres. Scale bar, 5 µm. Graph

comparing CENP-A SNAP and CENP-A fluorescence signal in in control and PHD2

depleted cells. Error bars indicate S.E.M. p values as in (B). (E) Alignment of CENP-

N proteins in higher mammals. The box highlights the putative hydroxylation site in

CENP-N. (F) HeLa Kyoto cells constitutively expressing GFP-CENP-N were bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 31

transfected with the indicated siRNAs for 72h. Scale bar 5 µm. Arrow points at

centromeric GFP-CENP-N. Graph comparing GFP-CENP-N fluorescence signal in in

control and PHD2 depleted cells at centromeres. Error bars indicate S.E.M. p-values

as in (B).

Figure 2: LC-MS analysis of in vitro hydroxylation of the CENP-N synthetic

peptides SLAPAALVCR by PHD2.

(A) ES-MS spectrum of the CENP-N peptide SLAPAALVCR at m/z 500.7835 (PHD2

substrate). (B) ES-MS spectrum of the product of in-vitro hydroxylation of the

synthetic peptide SLAPAALVCR showing a m/z increment of 7.9965 Th (mass

increment of 15.996 Da) corresponding to proline hydroxylation of the doubly

charged ion at m/z 500.7835 Th giving rise to ion at 508.7810 Th (the mass of the

hydroxylated peptide). (C) MS-MS spectrum of non hydroxylated peptide

SLAPAALVCR. (D) MS-MS spectrum of the hydroxylated peptide

SLAP(OH)AALVCR showing near complete y fragment ion series; fragment ions y7

and y8 have a mass increment of 15.996 Da confirming the site of hydroxylation as P

compared to the MS-MS spectrum of the non hydroxylated peptide shown in C. (E)

LC-MS analysis of CENP-N pinpointing the site of hydroxylation, CENP-N was

detected in the proteomics experiments with a sequence coverage of 18.98%.(F) XIC

of the in vivo hydroxylated peptide SLAP(OH)AALVcR where c is

carbamidomethylated cysteine at m/z 537.2924 and the non hydroxylated peptide

SLAPAALVcR at m/z 529.2953. The elution times are different because the gradient bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 32

was steeper. Mass differences (+ m/z 57) compared to synthetic peptides are due to

alkylation of the tryptic peptides.

Figure 3 PHD2 is required for CENP-N loading

(A) Alignment of CENP-N proteins in higher mammals. Arrows indicate the mutations

introduced in CENP-N. (B) Nuclear extracts of cells expressing GFP-CENP-N and

GFP-CENP-N mutants were immunoprecipitated with GFP-binder.

Immunoprecipitates were separated by SDS-page and analysed by western blotting

using the indicated antibodies. (C) U2OS cells expressing wt GFP-CENP-N or the

GFP-CENP-NA310G or GFP-CENP-NE304A/S308A (green) were stained for CENP-A

(red). Scale bar 5 µm. Graph comparing normalized fluorescence intensity of GFP in

GFP -CENP-N and mutant GFP-CENP-N expressing cells. Error bars indicate S.E.M.

p values are significant according to students test *** p<0.0001. >20 cells and >500

centromeres were quantified per condition. (D) PHD2 was depleted by siRNA

treatment of HeLa Kyoto cells for 72 hr before cells were stained for Dsn1 (red),

CREST (green) and DNA (blue). Scale bar 5 µm. Observed relative signal intensity is

shown in the graph on the bottom. Error bars indicate S.E.M. p values as in (A).

PHD2 depleted cells were stained for Hec1 (red) (E) and CENP-Q (red) (F). Cells

were costained with CREST (green) and DNA (blue). Scale bar 5 µm. Observed

relative signal intensity is shown in the graph on the bottom. Error bars indicate

S.E.M. p as in (A)

Figure 4 Hydroxylation of CENP-N increases binding to CENP-L bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 33

(A) Nuclear extracts from cell lines expressing GFP–CENP-N depleted with PHD2 or

CENP-N siRNA were separated by SDS–PAGE and analysed by western blotting

using the indicated antibodies. (B) Cell lines expressing GFP–CENP-N were exposed

to hypoxia for indicated time points. Extracts from treated cells were separated by

SDS–PAGE and analyzed by western blotting using the indicated antibodies. (C)

Autoradiograph of CENP-L CENP-N Co-immunoprecipitation. 6×Myc-CENP-L,

CENP-N and CENP-N P311A were expressed and labelled with 35S methionine and

the indicated proteins were mixed at equal stoichiometry. Each mixture was

immunoprecipitated with anti-Myc antibodies. Immunoprecipitates were loaded on

SDS-PAGE and analysed by autoradiography. Quantification of autoradiogram

signals is shown on the bottom. Bars represent mean values of three independent

experiments. Error bars represent S.D. p value significant according to student’s test

*p < 0.05, ** p < 0.01) (D) Immunofluorescence of cells expressing 6xmyc tagged

CENP-L treated with control, PHD2 or CENP-N siRNA for 72hr. Treated cells were

stained with anti-myc tag antibody (red) and CREST to mark kinetochores (green).

Scale bar 5 µm. Observed normalized signal intensity is shown in the graph on the

bottom. Error bars indicate S.E.M. p values are significant according to Students t-

test *** p<0.0001. 3 replicates, in each replicate >20 cells and >500 centromeres

were quantified per condition.3 replicates, in each replicate >20 cells and >500

centromeres were quantified per condition. (E) PHD2 was depleted by siRNA

treatment of HeLa Kyoto cells for 72 h before cells were stained for CENP-C (red),

CENP-A (green) and DNA (blue). Scale bar 5 µm. Observed relative signal intensity

is shown in the graph on the bottom. Error bars indicate S.E.M. p values as in (C). (F)

Immunofluorescence of cells expressing GFP-CENP-N and the CENPN mutant bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 34

A310G treated with CENP-N UTR siRNA for 72 h. Treated cells were stained with

anti-CENP-C antibody (red) and CREST to mark kinetochores (green). Scale bar 5

µm. Observed relative signal intensity is shown in the graph on the bottom. Error bars

indicate S.E.M. p values as in (C).

Figure S1 Depletion of PHD2 affects mitotic progression

(A) U2OS cells were transfected with siRNA against PHD2 for 72 h. After that cells

were fixed and stained with DAPI and the percentage of mitotic figures was

determined. Error bar indicate S.D. from three independent experiments (200 mitotic

cells were counted for each condition in each experiment) (B) PHD2 was depleted by

siRNA treatment of HeLa Kyoto cells for 72 h before cells were stained for CREST

(red), α-tubulin (green) and DNA (blue). Scale bar 5 µm. (C) U2OS cells were

seeded at low density and PHD2 was depleted by siRNA treatment for 96 h before

cells were stained for CENP-A (red) and DNA (blue). Scale bar 5 µm. Graph

comparing CENP-A fluorescence signal in interphase and mitotic cells in control and

PHD2 depleted cells. Error bars depict S.E.M. Three replicates, in each replicate >20

cells and >500 centromeres were quantified per condition. p values are significant

according to Students t-test *** p<0.0001. (D) PHD1 was depleted by siRNA

treatment of HeLa Kyoto cells for 72 h before cells were stained for CENP-A (red)

and DNA (blue). Scale bar 5 µm. Graph comparing CENP-A fluorescence control and

PHD2 depleted cells. Error bars depict S.E.M. Three replicates, in each replicate >20

cells and >500 centromeres were quantified. p values as in (C). (E) Experimental

scheme for SNAP based CENP-A assembly (E) and maintenance (F) assay

depicting order of synchronization, transfection and SNAP labeling steps described in bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 35

detail in Material and Methods. (G) GFP-CENP-N (green) expressing cells were

treated with DMOG for 48h and fixed and stained for CREST (red) and DNA (blue).

Scale bar 5 µm. Graph comparing GFP-CENP-N fluorescence in control and DMOG

treated cells. Error bars depict S.E.M. p values as in (C). (H) GFP-CENP-N (green)

expressing cells were depleted of PHD2 for 72h and fixed and labelled DNA (blue).

Scale bar 5 µm. Graph comparing GFP-CENP-N fluorescence in control and PHD2

siRNA treated cells in G1-, S- and G2-Phase. Error bars depict S.E.M. Three

replicates, in each replicate >20 cells and >500 centromeres were quantified. p

values as in (C).

Figure S2: Quantification of the stoichiometry of hydroxylation of CENP-N on

proline P311

(A) LC chromatogram showing the elution profiles of the synthetic peptides

SLAP(OH)AALVCR and SLAPAALVCR, the hydroxylated peptide at m/z 508.7829

elutes earlier (retention time (RT) = 51.50 min) than the non-hydroxylated peptide at

m/z 500.77851 (RT = 67.36 min); (B) Extracted ion chromatogram (XIC) of the

hydroxylated peptide (top) and non-hydroxylated (C) injected in increasing amounts

(1, 5, 10, 25, 50 ng) (D) Scatter plot showing the correlation between measured

peak areas for hydroxylated and non-hydroxylated peptides and amount of peptide

injected on RP-LC column; (E) Linear correlation between peak area of hydroxylated

peptide SLAP(OH)AALVCR vs peak area of non hydroxylated peptide

SLAPAALVCR.

Figure S3 CENPN levels are unaltered after PHD2 depletion bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Moser et al., PHD2 modulates CENP-N Function, p. 36

(A) Nuclear extracts from cell lines expressing GFP–CENP-N and cell lines

expressing GFP-CENP-N mutants were separated by SDS–PAGE and analyzed by

western blotting using the indicated antibodies. * marking a non specific band.

(B) Immunofluorescence of cells expressing GFP-CENP-N and the GFP-CENP-N

A310G (green) treated with CENP-N UTR siRNA for 96 h. Treated cells were stained

with anti-CENP-A antibody (red). Scale bar 5 µm. Observed relative signal intensity is

shown in the graph on the bottom. Error bars indicate S.E.M. p values are significant

according to Students t-test *** p<0.0001. 3 replicates, in each replicate >20 cells

and >500 centromeres were quantified per condition (C) U2OS cells transiently

expressing GFP-CENP-N or the mutant GFP-CENP-N E304A S308A (green). DNA is

stined with DAPI (blue). Scale bar 5 µm. (D). PHD2 was depleted by siRNA

treatment of HeLa Kyoto cells for 72 hr before cells were stained for CENP-K (red),

CREST (green) and DNA (blue). Scale bar 5 µm. Observed relative signal intensity is

shown in the graph on the bottom. Error bars indicate S.E.M. p values as in (A). (E)

U2OS cells expressing 6xmyc CENP-L were treated with control or PHD2 siRNA,

fixed and stained with anti-myc antibody (green) and DAPI (blue). Scale bar 5 µm.

Graph comparing CENP-L fluorescence signal in S- and G2-Phase. Error bars

indicate S.E.M. p values as in (A). (F) HeLa Kyoto cells were treated with control or

PHD2 siRNA, fixed and stained with CENP-C antibody (red), CREST (green) and

DAPI (blue). Scale bar 5 µm. Graph comparing CENP-C fluorescence signal. Error

bars indicate S.E.M. p values as in (A).

Table S1 Quantification of synthetic standards

Hydroxylated and non-hydroxylated peptides were injected on a reverse phase

column in increasing amounts (1, 5, 10, 25, 50, 100, 200 ng) and analysed using a

short gradient. Peak areas of hydroxylated and non hydroxylated peptides were

measured and correlated to peptide amount. Peptide areas were measured in

triplicates. Standard deviation (SD) and standard error (SE) were determined.

bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which 70was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Interphase Mitosis A 60 B

50 Prophase 40 si Control Prometaphase 30 Metaphase unaligned 20 Anaphase percentage of mitotic cells 10 si PHD2 0 siPHD2 siPHD2 siPHD2 si control DAPI siRNA1 siRNA2 siRNA2 +PHD2 GFP CENP-A C CENP-A SNAP:TMR CENP-A + DAPI 1.2

1.0 si Control 0.8 si Control

A intensity P- A 0.6 si PHD2

0.4 *** *** si PHD2 0.2 relative CEN

0 Interphase Mitosis 1.4 *** 1.2 E interphase 1 CENP-N

0.8 si Control Homo sapiens 0.6 si PHD2 Macaca mulatta Bos taurus 0.4 Canis familiaris *** Equus caballus 0.2

relative fluorescence intensity 0 CENP-A SNAP: TMR total CENP-A D F si Control si PHD2 CENP-A SNAP:TMR CENP-A + DAPI

CENP-N GFP si Control

si PHD2 + DAPI

1.2 1.2 1 1 0.8 si Control 0.8 0.6 si PHD2 0.6 0.4 *** *** 0.4 0.2

relative fluorescence intensity 0.2 0 *** CENP-A SNAP:TMR total CENP-A relative fluorescence intensity 0 *** Figure 1 si control si PHD2 Moser et al., 2015 bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is madey y y y y y y A available under aCC-BY 4.0C International license. 8 7 6 5 4 3 2 SLAPAALVCR 729.4068 100 b2 a2 y S L A P A A L V C R 201.1232 7 500.7835 90 173.1283 100 [y +2H]2+ b b 80 7 2 3 90 70 365.2071 80 y

ecnadnubA evitaleR ecnadnubA 6 70 60 y 632.3544 y8 60 5 501.2846 50 561.3168 50 y 800.4436 40 b 3 40 Abundance Relative 3 377.1955 30 272.1600 30 y4 20 20 501.7853 490.2806 10 10 0 500 501 502 503 504 505 506 507 508 509 510 100 200 300 400 500 600 700 800 900 1000

y y y y y y8 y7 6 5 4 3 2

S L A P(OH) A A L V C R B D b b SLAP(OH)AALVCR y 2 3 a 7 2 745.4011 508.7810 100 173.1283 100 90 b2 90 80 201.1231 2+ [y7+2H] 80 70 2+

ecnadnubA evitaleR ecnadnubA [b +2H] 373.2041 70 6 y 60 256.1295 8 60 y3 509.2823 50 AL-28 y5 50 b3 377.1961 816.4387 40 157.0971 272.1596 561.3152 y6 40 y4 30 y 30 Abundance Relative 2 490.2814 632.3541 20 278.1270 2+ 20 509.7830 [y8+2H] 10 10 408.7227 0 200 300 400 501 502 503 504 505 506 507 508 509 510 100 500 600 700 800 900 1000 m/z

E F RT 29.77 MA 325427 1 100 MDETVAEFIK RTILKIPMNE LTTILKAWDF LSENQLQTVN FRQRKESVVQ SLAP(OH)AALVcR 51 HLIHLCEEKR ASISDAALLD IIYMQFHQHQ KVWEVFQMSK GPGEDVDLFD 80 101 MKQFKNSFKK ILQRALKNVT VSFRETEENA VWIRIAWGTQ YTKPNQYKPT XIC of m/z 537.2924 151 YVVYYSQTPY AFTSSSMLRR NTPLLGQALT IASKHHQIVK MDLRSRYLDS 60 201 LKAIVFKQYN QTFETHNSTT PLQERSLGLD INMDSRIIHE NIVEKERVQR 40 251 ITQETFGDYP QPQLEFAQYK LETKFKSGLN GSILAEREEP LRCLIKFSSP 20 301 HLLEALKSLA PAALVCRIQK LLCYSGSHSQ GTQDPSSWQK DLYLLFVPLY Abundance Relative 351 PRC 27 28 29 30 31 32 33 34 P = Hydroxylated proline (HyP) RT 42.57 MA 1263163 100 SLAP AALV cR XIC of m/z 529.2953 80 60 40 20 R ti e Abundan e

40 41 42 43 44 45

ime (min)

Figure 2 Moser et al., 2015 A310G bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which CENP-N E304A,S308A P311A A was not certified by peer review) is the author/funder, who has granted DbioRxiv a license toDsn1 display the preprintCREST in perpetuity. +It DAPIis made available under aCC-BY 4.0 International license. Homo sapiens Macaca mulatta Bos taurus si control Canis familiaris Equus caballus B Input IP GFP-CENP-N GFP-CENP-N si PHD2

1.4 1.2 *** 1 control control A310G A310G wild type S310A E310A wild type S310A E310A 0.8 0.6 α-GFP 0.4 0.2

relative fluorescence intensity 0 α-PHD2 si control si PHD2

α-PHD2 E high exposure Hec1 CREST + DAPI

α-H3 si control

CENP-N GFP si PHD2 Wild type A310G E304A S308A

CENP-N GFP 1.2 1 0.8 0.6 *** 0.4 CENP-A 0.2

relative fluorescence intensity 0 si control si PHD2

F CENP-Q CREST + DAPI + DAPI

si control

1.2

si PHD2 0.8

0.6 *** 0.4 1.2

0.2 1 CENP-N fluorescence intensity 0.8 0 Wild type A310G E304A S308A 0.6 CENP-N GFP 0.4 *** 0.2

relative fluorescence intensity 0 si control si PHD2 Figure 3 Moser et al, 2015 bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to displayInput the preprint in perpetuity.anti-myc It is made IP A B available under aCC-BY 4.0 InternationalC license. + + + + + + + + CENP-L + - + - + - + - CENP-N - + - + - + - + CENP-N P311A 0 2 4 hrs of Hypoxia - - + + - - + + PHD2 si Controlsi PHD2 si CENP-N HIF1 α-CENP-N 6xmyc CENP-L CENP-N GFP CENP-N α-CENP-L * CENP-A α-Histone H4

Actin 1.8 * 1.6 1.4 1.2 1 D CENP-L 6xmyc CREST + DAPI 0.8 ** ** 0.6 0.4 0.2 normalised ratio CENP-N/ CENP-L normalised ratio CENP-N/ CENP-L si Control 0 CENP-N CENP-N P311A CENP-N CENP-N P311A +PHD2 +PHD2 E CENP-C CENP-A + DAPI si PHD2

si Control

si CENP-N

si PHD2

2 1.8 *** 1.6 1.4 2 1.2 1.8 *** 1 1.6 0.8 1.4 0.6 1.2 0.4 0.2 1 relative CENP-L intensity relative CENP-L 0 0.8 si Control si PHD2 0.6 0.4

F relative CENP-C intensity 0.2 0 GFP CENP-C + DAPI si Control si PHD2

Wild type CENP-N GFP

A310G

1.6 *** 1.4 1.2 1 0.8 0.6 0.4 0.2 relative CENP-C intensity 0 Wild Type A310G CENP-N

Figure 4 Moser et al., 2015 bioRxiv50 preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which A was not45 certified by peer review) is the author/funder, who has grantedB bioRxiv asi license Control to display the preprintsi PHD2 in perpetuity. It is madesi PHD2 40 available under aCC-BY 4.0 International license. 35 Prophase

30 Prometaphase 25 Metaphase 20 unaligned 15 Anaphase percentage of mitotic cells 10 5 0 DAPI si Control si PHD2 α-Tubulin si Control si PHD1 CREST C Interphase Mitosis D

si Control CENP-A

si PHD2 + DAPI

DAPI CENP-A 1.2 1.2

1 1

0.8 si Control 0.8 si PHD2 0.6 0.6 *** 0.4 *** *** 0.4

A intensity relative CEN P- A 0.2 0.2 relative CENP-A intensity relative CENP-A 0 0 Interphase Mitosis si Control si PHD1 E G CENP-N GFP CREST + DAPI 2nd thymidine 1st thymidine release release TMR -Star BTP (quench) pulse Control CENP-A assembly S G2 M G1 S G2 M G1 S

7.5h release synthesis of transfect nascent CENP-A pool quantify siRNAs CENP-A assembly DMOG

F TMR -Star pulse

-24h 0h 24h 48h 72h 1.2

1 quantify transfect CENP-A 0.8 siRNAs

0.6 0.4 *** H 0.2 CENP-N CENP-N CENP-N relative fluorescence intensity 0 G1-Phase S-Phase G2-Phase Control DMOG

1.2

si control 1

0.8

0.6 si control si PHD2 0.4 *** si PHD2 *** 0.2 relative fluorescence intensity *** 0 G1-Phase S-phase G2 Phase Supplementary Figure 1 Moser et al., 2015 bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November 3, 2015. The copyright holder for this preprint (which A was not certified by peer review) is the author/funder, who has granted bioRxivD a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. Non-hydroxylated peptide area Hydroxylated peptide area 2E+12

SLAP(OH)AALVCR 1.5E+12 RT 51.50 508.7829 100 SLAPAALVCR RT 67.36 1E+12 80 500.7851 60 Measured peak area 5E+11 40

Relative Abundance Relative 20

0E+00 0 50 100 150 200 48 50 52 54 56 58 60 62 64 66 68 70 72 74 76 Amount of peptide on column (ng) Time (min)

B E 2E+12 RT 66.47 y = 0.9275 x R2 = 0.98313

RT 66.81 1.5E+12 RT 66.78 RT 67.03 1E+12 RT 67.36 100 AA: 642122981585 50 ng 80 AA: 364342342245 25 ng 5E+11 60

AA: 262791352508 peptide for hydroxylated peak area Measured 40 10 ng AA: 68061109377 0E+00 0E+00 5E+11 1E+12 1.5E+12 2E+12 Relative Abundance Relative 20 5 ng AA: 12505920084 1 ng Measured peak area for non-hydroxylated peptide 50 55 60 65 70 75 80 Time (min) C RT 50.55 RT 51.66 RT 51.61 RT 51.35 RT 51.50 100 MA: 573876844866 80 50 ng AA: 13450998192 60 25 ng

AA: 14715168127 40 10 ng AA: 90386446281 Relative Abundance Relative 20 5 ng AA: 19281897224 0 1 ng 40 44 48 52 56 60 64

Supplementary Figure 2

Moser et al., 2015 bioRxiv preprint doi: https://doi.org/10.1101/030452; this version posted November GFP3, 2015. The copyrightCENP-A holder for this preprint+ DAPI (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made A available under aCC-BYB 4.0 International license. GFP CENP-N E304A WT A310G S308A P311A CENP-N WT α-GFP *

α- H4 CENP-N A310G C Wildtype E304A S308A 1.2

CENP-N GFP 1

0.8 ***

0.6

+DAPI 0.4

0.2

relative CENP-A fluorescence relative CENP-A 0 Wild type A310G D CENP-K CREST + DAPI CENP-N GFP

1.2

si control 1

0.8

0.6 ***

0.4 si PHD2 0.2 relative fluorescence intensity 0 si control si PHD2

CENP-L CENP-L E S-Phase G2-Phase 2.5

*** si control 2 *** 1.5 si control 1 si PHD2

si PHD2 0.5 relative CENP-L fluorescence relative CENP-L

0 S-phase G2-Phase F CENP-C CREST + DAPI 2 1.8 *** 1.6 si control 1.4 1.2 1 0.8 0.6 0.4

si PHD2 relative fluorescence intensity 0.2 0 si control si PHD2

Supplementary Figure S3 Moser et al., 2015